PRESR: a validated disease prediction tool for missende variants in STXBP1

PRESR is a machine learning algorithm that uses both sequence- and 3D structure-based features developed to improve pathogenicity prediction using 231 known disease-associated variants and validated using experimental data in vitro and in living neurons. PRESR outperformed existing tools substantially: Matthews correlation coefficient = 0.71 versus <0.55. PRESR was generated by Gurdeep Singh in Robert B. Russell lab Biochemistry Centre and BioQuant, Heidelberg University in collaboration with the Verhage lab @CNCR in Amsterdam and the Söllner lab @Biochemistry Centre, Heidelberg University, see PMID: 38490366 for details.